TgpuComputePipeline

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:98

Extends

TgpuNamable.SelfResolvable.Timeable

Properties

[$internal]

readonly [$internal]: ComputePipelineInternals

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:99

Overrides

SelfResolvable.[$internal]

[$soul]

readonly [$soul]: TgpuComputePipelineSoul

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:100

resourceType

readonly resourceType: "compute-pipeline"

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:101

Methods

[$resolve]()

[$resolve](ctx): ResolvedSnippet

Defined in: packages/typegpu/src/types.ts:435

Parameters

ctx

ResolutionCtx

Returns

ResolvedSnippet

Inherited from

SelfResolvable.[$resolve]

$name()

$name(label): this

Defined in: packages/typegpu/src/shared/meta.ts:44

Parameters

label

string

Returns

this

Inherited from

TgpuNamable.$name

dispatchWorkgroups()

dispatchWorkgroups(x, y?, z?): void

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:144

Parameters

x

number

y?

number

z?

number

Returns

void

dispatchWorkgroupsIndirect()

dispatchWorkgroupsIndirect<T>(indirectBuffer, start?): void

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:166

Dispatches compute workgroups using parameters read from a buffer. The buffer must contain 3 consecutive u32 values (x, y, z workgroup counts). To get the correct offset within complex data structures, use d.memoryLayoutOf(...).

Type Parameters

T

T extends AnyWgslData

Parameters

indirectBuffer

Buffer marked with ‘indirect’ usage containing dispatch parameters or raw GPUBuffer

GPUBuffer | TgpuBuffer<T> & IndirectFlag

start?

PrimitiveOffsetInfo pointing to the first dispatch parameter. If not provided, starts at offset 0. To obtain safe offsets, use d.memoryLayoutOf(...).

number | PrimitiveOffsetInfo

Returns

void

initAsync()

initAsync(): Promise<void>

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:150

Immediately resolves the pipeline, then awaits device.createComputePipelineAsync(). NOTE: it is not necessary to initialize pipelines manually.

Returns

Promise<void>

initSync()

initSync(): void

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:156

Immediately resolves the pipeline and creates WebGPU resources. NOTE: it is not necessary to initialize pipelines manually.

Returns

void

pipe()

pipe<T>(transform): T

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:142

Applies a transform to this pipeline, letting packages hand out reusable configuration steps, e.g. pipeline.pipe(cache.inject()).

Type Parameters

T

T

Parameters

transform

(pipeline) => T

Returns

T

toString()

toString(): string

Defined in: packages/typegpu/src/types.ts:436

Returns

string

Inherited from

SelfResolvable.toString

with()

Call Signature

with<Entries>(bindGroupLayout, bindGroup): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:107

Type Parameters

Entries

Entries extends Record<string, TgpuLayoutEntry | null>

Parameters

bindGroupLayout

TgpuBindGroupLayout<Entries>

bindGroup

TgpuBindGroup<Entries>

Returns

this

Call Signature

with(bindGroupLayout, bindGroup): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:111

Parameters

bindGroupLayout

TgpuBindGroupLayout

bindGroup

GPUBindGroup

Returns

this

Call Signature

with(bindGroup): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:112

Parameters

bindGroup

TgpuBindGroup

Returns

this

Call Signature

with<T>(immediate, value): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:122

Provides a value for the given immediate variable, applied on dispatch like the rest of the pipeline-held state and overridden by pass.setImmediates. The value is captured (copied) at call time; mutating it afterwards has no effect.

Passing an ArrayBuffer or typed array skips serialization entirely; the bytes are copied verbatim and the caller guarantees they match the schema’s layout.

Type Parameters

T

T extends AnyWgslData

Parameters

immediate

TgpuImmediateVar<T>

value

ArrayBuffer | ArrayBufferView<ArrayBufferLike> | InferInput<T>

Returns

this

Call Signature

with(pass): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:130

Directs subsequent dispatches into the given compute pass, letting multiple pipelines share one pass (and one submission).

Parameters

pass

TgpuComputePass

Returns

this

Call Signature

with(encoder): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:135

Directs subsequent dispatches into the given command encoder. Each dispatch records its own compute pass; the caller owns the submission.

Parameters

encoder

TgpuCommandEncoder

Returns

this

Call Signature

with(encoder): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:136

Parameters

encoder

GPUCommandEncoder

Returns

this

Call Signature

with(pass): this

Defined in: packages/typegpu/src/core/pipeline/computePipeline.ts:137

Parameters

pass

GPUComputePassEncoder

Returns

this

withPerformanceCallback()

withPerformanceCallback(callback): this

Defined in: packages/typegpu/src/core/pipeline/timeable.ts:13

Attaches a callback reporting the GPU-side start and end timestamps of the pipeline’s pass. Repeated executions within one command encoder write to the same query set indices, so the callback fires once, with the timestamps of the last execution.

Parameters

callback

(start, end) => void | Promise<void>

Returns

this

Inherited from

Timeable.withPerformanceCallback

withTimestampWrites()

withTimestampWrites(options): this

Defined in: packages/typegpu/src/core/pipeline/timeable.ts:15

Parameters

options

beginningOfPassWriteIndex?

number

endOfPassWriteIndex?

number

querySet

GPUQuerySet | TgpuQuerySet<"timestamp">

Returns

this

Inherited from

Timeable.withTimestampWrites